Warming speeds up range expansion in an experimental model system

Dispersal is becoming increasingly critical to understand as climate change is forcing species to shift their ranges to track changing environments. While we know that warmer temperatures can prompt range shifts, there is little known about how temperature influences the capacity for organisms to move by changing rates of range expansion. Warmer temperatures could affect the range expansion rate through two pathways: by increasing random, density-independent movement via increased metabolic rates, and/or by increasing population growth rates and driving density-dependent movement. Surprisingly, there have been no experimental tests of the effect of temperature on range expansion rate. We used red flour beetles, Tribolium castaneum, to test the overall effect of temperature on the rate of range expansion and to determine if random movement, population density, or both could be driving this effect. We grew beetle populations in linear connected landscapes at three temperatures (27.5 °C, 30 °C, and 32.5 °C) and tracked their expansion across these landscapes for 18 weeks. We then conducted separate assays to isolate the effects temperature on movement (density-independent dispersal probability) and on population growth rates. We found a positive effect of temperature on range expansion rate, with beetles in the 32.5 °C exhibiting the fastest range expansion. We also found positive effects of temperature on both dispersal probability and on population growth rate, indicating that both processes are likely contributing to this overall effect. Our findings highlight the importance of integrating the effects of temperature on range expansion speed in order to fully understand how quickly, species’ ranges could shift under climate change. 

Methods
Main experiment

We tested the effect of temperature on the rate of range expansion in Tribolium castaneum flour beetles across experimental linear landscapes in three temperature treatments. Landscapes were made up of 12 acrylic boxes (4 cm x 4 cm x 6 cm), where each box represented an individual habitat patch. We filled each patch with 30 g of a mix of a high-quality and a low-quality resource for this species (60 % whole wheat flour and 40 % rice flour). We used this mixed flour medium in order to temper population growth and make tracking population sizes logistically feasible. Patches were arranged in a single row and held together with elastic bands. Each patch had a 2-mm diameter hole drilled beneath the flour level on each side to allow for dispersal between patches.

We established ten replicate landscapes at each of three temperatures: 27.5 °C, 30 °C, and 32.5 °C. The historic temperature of our laboratory strain was 30C and the thermal optimum for this species is 30-31 °C. Landscapes were maintained in temperature- and humidity- controlled incubators (treatment temperature ± 0.5 °C, 50 %-70 % relative humidity). To control for any variation across incubators, we swapped treatments across chambers each month.

Each landscape was established by introducing 20 adult T. castaneum beetles into the first patch. Landscapes were then placed in their respective incubators for a three-week period to allow beetles to create tunnels in which to move through the flour. During this no-dispersal phase of the study, dispersal holes were blocked with acetate strips. After three weeks, the acetate strips were removed for 12 hours to allow for controlled dispersal into neighboring patches. After 12 hours, the acetate strips were added back in place to block the holes, and the landscapes were disassembled for population censusing. To census populations, we first inspected each patch in each landscape for signs of beetle activity (i.e., tunnels in the flour). We then sifted the flour out of all patches with any sign of beetle activity as well as the next two seemingly empty patches (to ensure no beetle presence). For any patch that contained adult beetles or larvae, we counted all live and dead adult beetles and replaced half of the flour medium with fresh flour medium. Upon completion of the count, beetles were placed back into their respective patches and landscapes were reassembled in the same order, placed back in their respective incubators, and the three-week cycle started again. This was repeated six times, for a total experimental duration of 18 weeks.

Movement rate assay

To determine the direct effect of temperature on movement rates (independent of the effect of temperature on population growth rates), we conducted a separate experiment. The experimental set-up was similar to that of the main experiment (same three temperatures and same flour medium), however landscapes consisted of only 2 individual habitat patches (acrylic boxes) and there were 5 replicate landscapes for each temperature treatment (for a total of fifteen 2-patch landscapes). These landscapes were maintained in the same incubators as the landscapes used in the main experiment. To initiate this assay, we introduced 10 adults into the first patch in each landscape, where they were allowed three weeks to acclimate and form tunnels. Acetate strips were removed for 12 hours to permit dispersal into patch 2. Upon completion of the 12-hour dispersal period, adult beetles were censused in each patch. Only one 3-week (21 day) cycle was completed to prevent population growth, as it takes between 35 and 62 days for T. castaneum to complete its life cycle on this flour medium and across our three experimental temperatures (unpublished data). This assay thus removed the possibility of population growth, allowing us to isolate the direct effect of temperature on dispersal rate.

Population growth rate assay

To determine the direct effect of temperature on intrinsic population growth rates (r), we conducted a separate experiment in isolated populations (i.e., no movement across habitat patches). For this experiment, we used the three temperatures and the same flour medium as described above, and had three replicate populations per temperature. To initiate the experiment, we added 15 adult T. castaneum to 10 grams in a lidded plastic container. We then allowed populations to grow and expand and counted all live and dead adult beetles in each population every two weeks for 22 weeks. After each count, we added an additional 5 grams of the appropriate flour type in order to provide a continuous supply of fresh resources.